Routing the wires and pipes of a decarbonising world
Grzegorz Marecki spends his days on a problem most people never see. Before a single power line, gas pipeline or fibre cable gets built, someone has to decide where it goes - across whose land, around which rivers, over what terrain, at what cost. For decades that decision has been made slowly, by hand, with a handful of options sketched in AutoCAD and checked against Google Earth. Marecki thinks that is the wrong way to build the twenty-first century.
As co-founder and CEO of Continuum Industries, the Edinburgh company he started in 2018, he leads the team behind Optioneer - a platform that treats infrastructure routing as a search problem. Instead of a dozen hand-drawn options, Optioneer generates and evaluates millions, weighing engineering constraints, environmental impact, planning rules and cost all at once. The engine at its core is a set of evolutionary algorithms, a form of AI modelled loosely on natural selection, which lets promising routes survive and reproduce while weaker ones drop away.
The pitch to the industry is blunt: nine out of ten infrastructure projects run late and over budget, and a lot of that pain is decided in the earliest weeks, when the route is chosen. Continuum says Optioneer can compress that early-stage planning from roughly twelve months to about eight weeks, and shave a meaningful slice off both design and construction costs. For the operators trying to string together the power lines and hydrogen pipelines the energy transition demands, faster routing is not a nicety. It is the bottleneck.
That framing has drawn real money. In September 2023 the company raised a $10 million Series A led by Singular, with earlier backers Credo, Playfair and Techstart returning alongside angels who had built businesses like UiPath, Skyscanner and SSE. The round took total funding to around $15.5 million and set a plain goal: double the team and push into new markets.
I arm engineers and planners with tools to do a 10x better job on dozens of projects- Grzegorz Marecki
A hyperloop pod, a hard pivot, and a bigger idea
Marecki trained as a civil engineer because he wanted to build the big systems that hold modern life together - roads, railways, grids, cities. The reality of the profession frustrated him. The path he saw ahead meant spending two decades on small components of large projects before ever leading one. He worked as an engineer, including a spell at the consultancy AECOM and time at contractor Bowmer & Kirkland, and kept looking for leverage.
He found it, of all places, in a student society. In 2017 Adam Anyszewski started an Edinburgh hyperloop team, HYPED, and pulled in Marecki and Emil Hansen. The group built two working prototype pods. Hyperloop itself never made commercial sense, and Marecki is candid about that. But the exercise forced the team to design a completely new kind of infrastructure with no legacy assumptions - and in doing so, they realised the valuable thing was not the pod. It was the software that had designed it.
In June 2018, joined by a fourth co-founder, Matt Blythe, they launched Continuum Industries to turn that insight into a business that could work across every kind of linear infrastructure. The moonshot quietly shifted from making hyperloop real to accelerating the unglamorous, essential work of getting net-zero infrastructure planned and permitted.
The early years were lean. Marecki has spoken about months when the bank balance sat near £20, and about learning to sell - the team's commercial instincts, he jokes, were sharpened somewhere between rickshaw hustle and enterprise software. What carried them, in his telling, was less a single clever idea than the discipline of showing up and pushing the same problem forward, week after week.
A dozen options, or a million?
The core bet is simple to state and hard to build: an engineer working by hand can seriously compare maybe a dozen routes. Software that searches the design space can weigh millions - and then hand the best few back to the human. The bars below sketch the gap Continuum is trying to close.
Capture
Pull in geospatial and engineering detail - terrain, land, constraints, environment - as data rather than static drawings.
Evolve
Evolutionary algorithms breed and test millions of routing options against cost, engineering and planning criteria.
Decide
Teams compare the strongest options, run scenarios, and move from concept toward permit far faster.
On execution, and the long game
Ideas are often cheap and execution is what matters the most.
Just continuing to push forward consistently over time will help you get to places you wouldn't have thought were available.
Nine out of ten infrastructure projects have long been behind schedule and over budget.
I arm engineers and planners with tools to do a 10x better job on dozens of projects.
The goal is an operating system for infrastructure - from concept to permit, faster and cheaper than anyone thought possible- The vision behind Continuum Industries
Five things that stick
An engineer fixing his own trade
He is a civil engineer who ended up writing software to solve the frustrations he felt on the job.
Born in a student society
Continuum grew out of a hyperloop club, not a lab or a corporate spinout.
AI by natural selection
Optioneer's engine uses evolutionary algorithms - a "soft" AI that mimics survival of the fittest.
He hosts the conversation too
Marecki runs an interview series, "Tete-a-tete with Grzeg," with figures across the energy industry.
Rising Star
Founders Forum Europe named him one of its 2023 Rising Stars.